BackgroundSorghum (Sorghum bicolor L. Moench) productivity is severely impeded by low phosphorus (P) and aluminum (Al) toxic soils in sub-Saharan Africa and especially West Africa (WA). Improving productivity of this staple crop under these harsh conditions is crucial to improve food security and farmer’s incomes in WA.ResultsThis is the first study to examine the genetics underlying sorghum adaptation to phosphorus limitation in a wide range of WA growing conditions. A set of 187 diverse sorghum genotypes were grown in 29 –P and + P field experiments from 2006-2012 in three WA countries. Sorghum grain yield performance under –P and + P conditions was highly correlated (r = 0.85***). Significant genotype-by-phosphorus interaction was detected but with small magnitude compared to the genotype variance component. We observed high genetic diversity within our panel, with rapid linkage disequilibrium decay, confirming recent sequence based studies in sorghum. Using genome wide association mapping based on 220 934 SNPs we identified one genomic region on chromosome 3 that was highly associated to grain yield production. A major Al-tolerance gene in sorghum, SbMATE, was collocated in this region and SbMATE specific SNPs showed very high associations to grain yield production, especially under –P conditions, explaining up to 16% of the genotypic variance.ConclusionThe results suggest that SbMATE has a possible pleiotropic role in providing tolerance to two of the most serious abiotic stresses for sorghum in WA, Al toxicity and P deficiency. The identified SNPs can help accelerate breeding for increased sorghum productivity under unfavorable soil conditions and contribute to assuring food security in WA.Electronic supplementary materialThe online version of this article (doi:10.1186/s12870-014-0206-6) contains supplementary material, which is available to authorized users.

The first sorghum [Sorghum bicolor (L.) Moench] hybrids based on West African Guinea‐race‐derived parents were created to enhance farmer's food security and income through increased yields. To assess their performance, eight hybrids, six experimental pure‐line cultivars, one pure‐line check (Lata), and a highly adapted landrace cultivar (Tieble) were evaluated in 27 farmer‐managed and two on‐station yield trials in Mali, West Africa, from 2009 to 2011. The hybrids were confirmed to have photoperiod sensitivity similar to the well‐adapted Guinea landrace check cultivar. Genotypic differences for on‐farm grain yield were highly significant and genotype × environment crossover interactions were limited. The yield superiorities of individual hybrids, relative to the landrace check, ranged from 17 to 37% over the 27 on‐farm trials. The three top yielding hybrids showed 30% yield advantages across productivity levels, with absolute yield advantages averaging 380 kg ha−1 under lower (1.0–1.5 t ha−1) and 660 kg ha−1 under higher (2.0–3.5 t ha−1) productivity conditions. A mean male‐parent (better parent) heterosis of 26% was observed for the four hybrids having Lata as a male parent. As the hybrids studied here were obtained with a low intensity of selection using a limited number of parents, even greater yield superiorities may be attained with development of distinct parental pools and scaled‐up hybrid breeding.

Sorghum (Sorghum bicolor (L.) Moench) is widely cultivated in West Africa (WA) on soils with low phosphorus (P) availability. Large genetic variation for grain yield (GY) under low‐P conditions was observed among WA sorghum genotypes, but information is lacking on the usefulness of P‐tolerance ratios (relative performance in –P [no P fertilizer] vs. +P [with P fertilizer] conditions) and measures of P‐acquisition and internal P‐use efficiency as selection criteria for enhancing GY under low‐P conditions. We evaluated 70 WA sorghum genotypes for GY performance under −P and +P conditions for 5 yr in two locations in Mali and assessed P acquisition (e.g., P content in biomass) and P‐use efficiency (e.g., grain produced per unit P uptake) traits under −P and +P conditions in one site in 2010. Significant genetic variation existed for all P‐tolerance ratios across multiple sites. Photoperiod‐sensitive landrace genotypes showed significantly better P tolerance and less delay of heading under P‐limited conditions compared with photoperiod‐insensitive varieties. Genotypic correlations of P‐tolerance ratios to GY under −P were moderate. Phosphorous acquisition and P‐use efficiency traits independent of harvest index were of similar importance for GY under −P conditions in statistically independent trials. However grain‐P and stover‐P concentrations from one −P trial showed only weak correlations with GYs in statistically independent trials. Highest predicted gains for −P GY were obtained by theoretical index selection based on −P GY combined with P‐use efficiency traits (e.g., low‐grain P concentration). Such index selection is expected to achieve both increased sorghum productivity and P sustainability in the P‐limited WA production systems.

Many farmers in West and Central Africa (WCA) prefer tall (>3 m) grain sorghum [Sorghum bicolor (L.) Moench] for various reasons. This study seeks to determine (i) what yield superiority newly bred, tall, photoperiod‐sensitive guinea‐race sorghum hybrids can provide relative to an adapted landrace variety across a wide range of productivity conditions, and (ii) the risk of these hybrids failing to provide yield superiority for individual farmers. Seven hybrids, one local check, and eight pure‐line progenies were evaluated in 37 farmer‐managed, on‐farm yield trials across three Malian zones and 3 yr. Environments were classified into four productivity groups (low [0.78–1.10 Mg ha−1], mid‐low [1.10–1.50 Mg ha−1], mid‐high [1.50–2.00 Mg ha−1] and high [2.00–2.65 Mg ha−1]) based on their trial mean grain yield. Mean yields of the seven tall hybrids were 3 to 17% (ranging from 0.06 to 0.28 Mg ha−1) higher than that of the local check across all environments and were highest (14–47%) averaged across the seven trials with the lowest mean yields. The individual overall highest‐yielding hybrid showed superiorities over the local check in the low, mid‐low, mid‐high, and high productivity levels of 0.43 (47%), 0.14 (10%), 0.47 (27%), and 0.34 (14%) Mg ha−1, respectively. The tall hybrids rarely had yields significantly inferior to the local check. Farmers’ preference for, and the possible benefits of, taller plant types may lead farmers to grow tall hybrids, particularly under the typical low‐productivity production conditions of WCA.

Witchweed (Striga spp.) infestations are the greatest obstacle to sorghum [Sorghum bicolor (L.) Moench] grain production in many areas in Africa. The objective of this study was to evaluate the efficacy of herbicide seed treatments for controlling Striga infestation of sorghum. Seeds of an acetolactate synthase (ALS) herbicide–tolerant sorghum hybrid were treated with two ALS‐inhibiting herbicides. Treatments included three rates of imazapyr (IMI), three rates of metsulfuron‐methyl (MET), and an untreated control group. In greenhouse trials, observations at 32, 46, and 60 d after planting showed that seeds treated with the highest herbicide rates had the fewest Striga attachments and the greatest delay in attachment. All plants in the untreated group died at or before sorghum flowering; however, herbicide seed treatments, particularly metsulfuron, reduced Striga emergence and significantly increased sorghum grain yield and dry matter production. Field studies comparing seed treatments produced similar results with delayed Striga emergence and fewer emerged Striga plants in herbicide‐treated plots. These studies indicate that herbicide seed treatments may provide a highly effective tool for managing Striga in sorghum.

S orghum bicolor (L.) Moench is a major staple cereal crop in the Sahelian and savannah zones of West Africa (WA), where it constitutes the main cereal in the diet of millions, mostly living in rural areas. Yields on farmers' fields average only 100 g m −2 (1 t ha −1 ) (FAOSTAT, 2018), despite the development of new varieties (Yapi et al., 2000;Atokple, 2003). In recent years, farmers in the savannah zones of WA have replaced sorghum with early-maturing maize (Zea mays L.) varieties, especially in cotton (Gossypium hirsutum L.)-producing systems, where fertilizer is easily available (Staatz et al., 2011;Theriault et al., 2018). However, area under sorghum has been increasing in these countries (FAOSTAT, 2018), because sorghum is the cereal crop that can be grown successfully without the use of mineral fertilizer. Thus, new sorghum varieties can provide advantages to farmers, especially if they are well adapted to low soil fertility, especially low-P (LP) conditions. This is particularly important

Many efforts have been made to improve sorghum [Sorghum bicolor (L.) Moench] varieties, but adoption of improved varieties remains low. Sorghum has diverse panicle architecture and grain qualities that vary within and between races, and utilization and adoption may depend on these traits. Recent efforts in West Africa to improve local guinea race germplasm as a base material have diversified potential options: there are breeding materials with a range of panicle types with increased grain number per panicle and a range of droopiness, as well as laxness and threshability. This study was designed to expand our understanding about sorghum grain and panicle traits that are important for farmers in the Sudan savanna zone of Mali. We combined a sorghum panicle sorting activity with qualitative interviews in Mande and Dioïla to understand farmers' knowledge and preferences about sorghum characteristics. A total of 20 panicle sorting activities and 20 interviews were conducted with men and women sorghum producers. Based on their roles and responsibilities in sorghum production and processing, farmers associated specific panicle types, plant types, and grain traits with aspects of pest control, threshability, storage duration, and yield. Farmers preferred open panicles and droopy architecture for disease and pest control; hard grains for storage and appropriate ratios of flour and grits; high density of grain on the panicle for yield; and specific glume qualities for threshability. Breeding programs need to consider these regional preferences and gender roles to develop appropriate material and increase adoption of productive varieties.

Création variétale décentralisée Sélection participative des variétés de sorgho à l'aide d'essais multilocaux dans deux zones cibles